The overall goal of this collaborative project is an understanding of mechanisms involved in the humoral response created by a breast tumor in its human host and employing them for breast cancer therapy. The identification of B-cell clones participating in this response will allow the use of their genetic information to produce stale transfected cell lines secreting human monoclonal antibodies binding antigens on breast tumors. The identification of epitopes/antigens involved in the patient's immune recognition of the tumor will shed light on the mechanism by which humoral and cellular immune response can control tumor growth and spread. Information on the antigens' function can open new areas of research related to immune control, prevention, and therapy of breast cancer. The Specific Aims of this project are: 1. Define the epitope structures on the breast mucin tandem repeat domain that are the most effective targets for radioimmunotherapy and immune recognition, and determine factors involved in its altered processing in breast tumor cells. 2. Study the 46 kDa breast glycoprotein, identifying its epitopic domains that provide the best targets for therapy, and explore its possible paracrine/autocrine function. 3. Characterize the antigens and their epitopes recognized by human and F-v grafted MoAbs produced. Certain MoAbs against the breast mucin and the 46 kDa HMFG glycoprotein are effective in radioimmunotherapy. Because of the altered processing of the breast mucin some epitopes in the tandem repeat domain are preferentially expressed in breast carcinomas, and thus certain MoAbs recognizing these normally cryptic epitopes are more effective than others in imaging metastases in breast cancer patients. Also, in the case of the 46 kDa antigen, some MoAbs against it are much more effective than others in preclinical RIT studies. Our cDNA cloning studies show that the 46 kDa antigen has domains suggesting an auto- or paracrine function: a C-terminal domain containing cell attachment and EGF-like sequences and a N-terminal domain homologous to the phospolipid-binding C1 C2 regions of human coagulation factors V and VIII. The human MoAbs prepared in the other projects may identify epitopes on these latter two antigens, if so, the precise epitopes will be determined. If new antigens are identified other antigens their cDNAs will be cloned and sequenced. The epitopes will be analyzed by epitope mapping and production of recombinant antigens and their fragments as chimeric proteins. The hypothesis that epitope structural knowledge can predict effectiveness in RIT will be tested by production of new MoAbs using selectively constructed epitopic structures for immunization and selection. The therapeutic effectiveness will be tested preclinically.